A Mass-Gravity Scaling and Beam-/Shell-Volume Coupling Framework for Efficient Pile Installation in Centrifuge Tests




A Mass-Gravity Scaling and Beam-/Shell-Volume Coupling Framework for Efficient Pile Installation in Centrifuge Tests


Physical modelling, particularly through centrifuge testing, provides essential insights into geotechnical processes such as pile installation involving large deformations at realistic stress levels. However, high-fidelity numerical replication of these processes for comprehensive analysis is often computationally intensive. This paper introduces and validates a combined numerical approach designed to significantly improve the efficiency of explicit simulations for quasi-static soil-structure interactions. Explicit integration schemes are essential for large deformation analyses, yet the critical time increment is typically governed by finely meshed, non-rigid structural elements, leading to excessive simulation times. To address this, two techniques are combined: Mass-Gravity-Scaling (MGS), which increases the stable time increment by scaling density and gravity while preserving the initial geostatic stress state, and a kinematic beam-/shell-volume coupling method for the computationally efficient modelling of slender deformable bodies. Both techniques are implemented within the Coupled Eulerian-Lagrangian (CEL) framework and benchmarked against 50g centrifuge test data for jacked pile installation in sand. The combined methods achieve a computational speed-up of approximately 6 times at a moderate scaling factor (S = 10) and up to 13 times (S = 100) compared to the conventional approach, while maintaining high fidelity to the physical test results (mean percentage error remains below 7% for moderate scaling factors). The study provides a validated and practical pathway to make explicit simulations more efficient, enabling broader parametric studies and the integration of more advanced soil models within feasible runtimes.



D. Alkateeb; Jurgen Grabe


11th International Conference on Physical Modelling in Geotechnics (ICPMG2026)



Special Session 5: Combination of numerical and physical modelling



https://doi.org/10.53243/ICPMG2026-45